Does Spam In Google Analytics Result In Bad User Metrics For SERPS?

[seoskunk]

OK so I having to explain to some website owners why there bounce rate has increased and time on site decreased and when I look at Analytics I just don't get it.

Its full of referrers that don't link to the site. Hits that never existed and time on site 0.00.

Direct traffic is a joke I have one site showing 93% bounce rate on direct traffic. COME ON!

I compare to statcounter and it seems to filter out this shit, why can't google?

But the thing that worries me is if they can't filter out this in Analytics which has such a massive usage are they filtering this data for user metrics on search?

[goodroi]

Almost sounds like an old case of referral log spamming. I have heard of some people experimenting with sabotaging user metrics but that is very rare.

Have you seen any change in the Google traffic? Just because you see something in your analytics does not mean Google is measuring it and using it to influence your rankings.

[glakes]

Google used to do a good job at eliminating referrer spam from analytics. Whether it's a new form of referrer spam or Google's analytics team dropping the ball, too much referrer spam is now being calculated/counted in their stats. Considering some of the referrer hits logged in analytics is coming from notorious referrer spammers, I'd have to say Google is neglecting this problem.

[netmeg]

Google's not going to automatically remove it because if they did there were would be screams about yet more overreach. But there's lots of resources out there for filtering it out of your Analytics. (And by the way, MY StatCounter account most certainly doesn't filter it out) There are also a few suggestions for blocking in .htaccess or Apache - I've tried a few of these, but only had limited success with it.

[aristotle]

Does Spam In Google Analytics Result In Bad User Metrics For SERPS?

If Google has any sense at all, it won't use Analytics data for User Metrics. In fact the main reason why Google created its Chrome browser is to get user metrics data. The Chrome browser can collect far more-detailed and far more reliable data than Analytics can. Not to mention the fact that many sites don't have Analytics installed, whereas the Chrome browser can collect user metrics data for all sites, not just those that use Analytics.

In sum, Google would be very stupid to use Analytics data for user metrics, when they can get far better data with their Chrome browser. In fact, as I said, that's why they expended the time, money and effort to create the Chrome browser in the first place.

[netmeg]

I expect Google uses some of everything they got.

[glakes]

Google's not going to automatically remove it because if they did there were would be screams about yet more overreach.

An overreach by eliminating widely known referral spammers from reports? I don't think that's an overreach, it's about making their analytics service more useful out of the box. Since Google does not handle that on their end, millions of webmasters must create rules on their own. Maybe this is like Google's disavow tool, which relies on the actions of millions of webmasters to build a data set that can be used elsewhere. Or, on the other hand, the Google analytic team is asleep at the wheel.

[lucy24]

There are also a few suggestions for blocking in .htaccess or Apache - I've tried a few of these, but only had limited success with it.

You can only block visits to your own site. As I understand it, the current problem is with people* who proceed directly to analytics, which lives on google's servers. They don't even bother to visit the site itself, because they can achieve the same purpose without it. There's a longish recent thread-- which, naturally, I now can't find-- about referer spam in analytics.


* For, ahem, a given definition of "people".

[seoskunk]

You can only block visits to your own site. As I understand it, the current problem is with people* who proceed directly to analytics, which lives on google's servers. They don't even bother to visit the site itself, because they can achieve the same purpose without it. There's a longish recent thread-- which, naturally, I now can't find-- about referer spam in analytics.

Absolutely correct I got 40 "visits" from RU yesterday in Analytics with 100% bounce rate, they didn't even bother to set a referrer and on comparing the same site on statcounter visits from RU were 0. None of these were actual visits to the site. All came from Chrome browser, I see no motivation to do this unless its to destroy user metrics, its not even referrer spam anymore. I think google have dropped a ball here, perhaps using an 1px image to verify the visit like in the old days would sort this problem out overnight,